The internet is breaking. And it’s happening fast. Your favorite websites are going down, your favorite apps are not working, your favorite services are not available. And it happens all the time, at a rate that is concerning. There is a fix.
You know that Meta, Google, Apple, Amazon, Microsoft, etc. are the players owning all the big platforms on the internet. If one of them goes down, their platforms become unavailable. However, there are many more influential players that most people are not even aware of. If one of them goes down, half of the internet becomes unavailable.
Cloudflare and AWS are arguably the two most important players in the internet. They are the backbone of the internet as they provide the infrastructure that allows companies to run their services. They are the ones that keep the internet running. Cloudflare went down again this morning for a few minutes. AWS had a huge outage in October lasting several hours. These are the big players, ruling the internet. And they are breaking, at a rate that should get us thinking.
The internet is built on a foundation of trust. It relies heavily on a few big players to keep the internet running. Most people don’t know how the internet works anyways (which is fine). Some might think it’s a trustless system with so much competition that it’s not a problem. Indeed, there is a lot of competition, but the monopoly of the big players is still there. The internet is not a trustless system, it is a system of trust. And that trust is breaking as we see more and more outages.
The issue is not a lack of competition, it is a lack of resilience. There is a single point of failure, one company everyone is dependent on. Dependencies are bad from a technical perspective since a system’s resistence suffers when the dependency chain grows. The bigger the dependency chain, the more likely it is that a single point of failure will cause a cascading failure.
The same applies to one’s own data, we do backups to lower the risk of total loss of our data - unless you are a South Korean government service. Often the fix for this risk is redundancy, meaning replicating the object at risk, here the data or the platforms providing infrastructure. Instead of choosing one cloud provider (AWS), a company would use a secondary cloud as a backup solution (e.g. Azure). This often comes with hefty costs and maintenance overhead, however, so this is often not the most viable solution.
Spinning the idea of replication further, you at some point will reach the concept of decentralization. In the first place it just means the absence of a central entity, which is not true in our scenario, since we just “backup” central entities multiple times. There is no clear line between centralized and decentralized systems, although one could argue that the system design defines already whether something is centralized or not.
Since manually “decentralizing” it’s own infrastructure comes with several downsides (costs, maintenance, resources) from a company and individual perspective, the true innovation lies is decentralization by design. A technology built on this philosophy from the ground up, is the blockchain (I refer to the idea, rather than one specific implementation). The idea of a new, resilient, decentralized and thus more democratic internet has been born and is commonly referred to as Web3.
Long live the internet! Keep building!